One Spark should be powerful enough to run multiple chambers, it is not powerful enough to run a database and web server.
In the future, all control will not be centered around chambers, but around PIDs.
Each PID takes a setpoint and a sensor as input, but they are mostly independent.
One fermentor could be 3 PIDs: one for heating, one for cooling and 1 to give a fridge set point from the beer temp.
Then you could have another 3 PIDs for a second chamber and 3 for your mash setup.
But it could be much simpler: one sensor, one heater, one setpoint, one PID.
Multiple beers in a single fridge could be:
- 1 PID for beer 1 (heater)
- 1 PID for beer 2 (heater)
- 1 PID to cool the fridge
- 1 PID to set the fridge setpoint (based both beers? Need to think about how this would work exactly)
So by making it just a list of PIDs, this could be very flexible.